Mathematical-based morphological classification of skin eruptions corresponding to the pathophysiological state of chronic spontaneous urticaria

Background Chronic spontaneous urticaria (CSU) is one of the most intractable human-specific skin diseases. However, as no experimental animal model exists, the mechanism underlying disease pathogenesis in vivo remains unclear, making the establishment of a curative treatment challenging. Methods A novel approach combining mathematical modelling, in vitro experiments and clinical data analysis was used to infer the pathological state of CSU patients from geometric features of the skin eruptions. Results Based on our hierarchical mathematical modelling, the eruptions of CSU were classified into five categories, each with distinct histamine, basophils, mast cells and coagulation factors network signatures. The analysis of 105 real CSU patients with this classification by six individual dermatologists achieved 87.6% agreement. Furthermore, our network analysis revealed that the coagulation status likely determines boundary/area pattern of wheals, while the state of spontaneous histamine release from mast cells may contribute to the divergence of size and outline of the eruptions. Conclusions Our multi-faceted approach was accurate in defining pathophysiological states of disease based on geometric features offering the potential to improve the accuracy of CSU diagnosis and better management of the disease in the clinic.